Quote from: zeroth on June 28, 2017, 09:06:27 PM

Thank you for your replies. These have led me to a few more questions that I'd be really interested in discussing more.

8. (Sorry if this is a stupid question) Stepping back a little, why isn't blockchain technology scalable? Let's say one person makes one transaction per day paying a fee of 1 unit which is enough to attract enough miners to process the transactions in a practical time period. Won't a billion people each making one transaction per day pay a billion units in fees? So why isn't there a proportional incentive for miners and hence the number of miners scaling up accordingly?

9. Why does centralisation improve scalability?

10. How does market sharing between competitor networks or sidechains actually help? Surely you still have the same problem with a limited amount of processing power for, say, 100 transactions per second on one network versus 50 transactions per second on each of two networks?

11. Do we know any more details about these 3.0 platforms that Fuserleer mentioned? How are they fundamentally different?

8. Miners don't have anything to do with scalability...they just produce the blocks. The problem is the architecture of a block chain. Mainly the issue is that when blocks get past a certain size (say 1GB for arguments sake) it takes longer than 10 minutes for that block to make its way to all nodes in the network. By then ANOTHER block is produced, which also takes longer than 10 mins, then another, then another. All the nodes just end up collecting blocks and (potentially) reorganizing the chain. At this point it all grinds to an almost halt.

9. Say the Bitcoin network is 10,000 nodes, and another Bitcoin network is 10,000 nodes but has 100 master nodes. Its MUCH easier to keep 100 nodes in sync with each other than 10,000, so bigger blocks could be used for more TPS. The problem is that its also easier to attack 100 nodes and take them offline, or for 100 node operators to collude to abuse the network than it is for 10,000.

10. Having many networks / side chains can help, but this also brings some problems. First of all, you only have X amount of hashing power (or stake, or whatever). If you then have 100 side networks / chains, in most cases you have to share X. Each network then has the security equivalent to X/100 and its much easier for an attack to take control of that particular side network / side chain.

Another issue is that if you want to transact outside of your side network / chain, it is a complicated and slow process, so the benefits are only really seen if you are transacting with someone in the same network / chain as you.

11. Most of the 3.0 tech being developed is really early stage (other than my own project Radix). Theory mainly and little code hence the 18 month window. I'm not even sure if they are announced publicly yet, just discussions I've had with people at events and meetups with ideas. I'll check with the developers before I name them.